Metraipolis › archive › dc_d33d59fe72fb
Aligning Medical Vision Language Models at the Capability Boundary with Entropy Adaptive Rubric Grounding — arXiv:2610.39835
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.39835
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-31T06:13:25.925Z
- SHA-256 of the archived text
74ef0764dc3284622108075f54b4ea9531506f32e83047dc16d3c53c04696e3b- Stable URL of this page
https://metraipolis.com/documento/dc_d33d59fe72fb
How to cite
Aligning Medical Vision Language Models at the Capability Boundary with Entropy Adaptive Rubric Grounding — arXiv:2610.39835 — arXiv (en), https://arxiv.org/abs/2610.39835 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d33d59fe72fb (Metraipolis archive dc_d33d59fe72fb, sha256 74ef0764dc3284622108075f54b4ea9531506f32e83047dc16d3c53c04696e3b)
BibTeX
@misc{dcd33d59fe72fb,
title = {Aligning Medical Vision Language Models at the Capability Boundary with Entropy Adaptive Rubric Grounding — arXiv:2610.39835},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.39835},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d33d59fe72fb; sha256 74ef0764dc3284622108075f54b4ea9531506f32e83047dc16d3c53c04696e3b},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.39835. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Medical diagnostic reasoning requires verifiable links between visual evidence and clinical knowledge, yet current medical vision language models are aligned with objectives that reward the wrong behavior. Uniform supervised fine tuning lets easy cases dominate updates while clinically critical boundary cases saturate prematurely. Sparse rewards based only on final-answer correctness cannot assess anatomical grounding or clinical logic, inviting reward hacking through answer shortcuts. We propose MedCBR, a two-stage alignment framework coupling boundary-aware preference optimization with rubric-grounded reinforcement learning. First, we construct counterfactual boundary preferences at the model's capability frontier using empirical pass rates and preference pairs that share reasoning prefixes but diverge at decisive diagnostic branches. Difficulty Weighted Boundary Preference Optimization (DBPO) concentrates learning on this calibrated boundary region. Second, we distill reusable clinical rubrics from boundary trajectories and activate them according to prediction entropy. Entropy Gated Grounded Policy Optimization (EGPO) rewards answer correctness, anatomical grounding, and reasoning fidelity while suppressing rubric noise on reliable samples. A reference policy anchored to the boundary-aligned model couples population-level boundary priors with instance-level uncertainty, making diagnostic reasoning more accurate and clinically grounded. Experiments on six medical benchmarks show consistent improvements over strong general, specialized, and reinforcement learning baselines, reaching 61.2% average accuracy with the largest gains on challenging differential diagnosis tasks. Ablations show super-additive gains: combining the two stages exceeds the sum of their individual improvements. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.39835 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-31T06:07:08.736Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).